Random Notes on AI and ML
Some interesting things from the world of Artificial Intelligence and Machine Learning -
In China, a CCTV camera caught a prominent business woman Dong Mingzhu jaywalking. The CCTV system, acting as a police and judge, displayed her image on a big public screen to name and shame her. Unfortunately, the CCTV camera had mistakenly caught the image of her on signage of a bus passing by. She was nowhere near the area!
Recently an algorithm has been created to ascertain whether a person has actually understood the text they have read or are just pretending to have. The algorithm, developed by a team at Northwestern University in Evanston, Illinois, US, receives and processes neuron signals emitted from the reader’s brain. Based on the pattern of the brain’s electrical signals, collected by some sensors, the algorithm is able, with an accuracy of 90%, to determine if the person who read the text really understood it.
First of its kind fraud using AI generated deepfake call has been reported. The CEO of a company in UK got a call from his boss, the CEO of the German parent company asking him to transfer 220,000 euros immediately to the bank account of a Hungarian supplier. The voice was similar to the German boss with his accent and melody. The money when transferred to the Hungarian account was moved to Mexico and then disbursed ahead to other locations. When the UK CEO received another call request, he became suspicious and the whole thing became clear. With deepfakes, what do you trust and whom do you trust?
And, what if the bots are reading this and get even better at it (while some of us are busy sharing Good Morning messages on What's App groups)?!